In accompany with the increasing of Internet popularization rate to all people, people get used to utilizing Internet to issue, obtain and transfer information day by day, the Internet has already become a main carrier of various information, and reflects the public subjective emotions to social reality to a certain extent. More often, there are a lot of websites reprinting and reporting hot issues over the Internet, it has become our new needs on how to use original machine resources more reasonable, and obtain more information in a certain time without increasing the processing performance of those original hardware resources.
Most of the present downloading models proceed downloading based on single task single-thread and multitask multi-threads, in which, when there is a task, a thread will start-up, then the information is received, obtained and returned. It is easier to understand and to operate such kind of models. When the task amount is not huge, the downloading can be well controlled by utilizing the thread(s), and it's easy to track the whole procedure for obtaining information. However, due to the fact that once a task is received, a thread needs to start-up to receive and perform the task, and there is a close relationship between the quantity of received tasks and the performance of the machine, one of the deficiencies of such model is consuming large amount of system resources, and having a higher requirement to the machine.